GFS test

[cycloneye]

http://www.nco.ncep.noaa.gov/pmb/nwprod ... loop.shtml

00z loop shows a hurricane off the east coast but as always need more confirmation and other model support but GFS has shown this for many of the 00z-12z runs.

[cycloneye]

http://www.nco.ncep.noaa.gov/pmb/nwprod ... loop.shtml

12z loop.It shows a weaker system not a hurricane but a weak storm north of Puerto Rico by late next week.Now let's see what the 00z run shows and if other models go with this.

[Derek Ortt]

one other caution about model runs


Once you get beyond 5-7 days, the model output is also worthless as the numerical equations being solved are subject to roundoff error. These errors grow exponentially with time; thus, anything passed 5-7 days has so much error in it that the output again has less value than the paper it is printed on

[cycloneye]

Yes agreed.Beyond 7 days is not good data although we can look at very long range to see if the model is consistant in a trend and when the time comes then we can see if they busted or not that long range.

[jimvb]

I view GFS and other models like driving down a highway, especially one where you can see a long distance. I look down the highway. I see a tractor-trailer way way off as a small dot on the horizon. It is in the other lane. Does this mean it will pass by me on the road? No. It could turn right onto an intersecting road before it gets to where I am, or it could pull over at a rest stop and disappear from the highway. But still the appearance of this truck indicates that about a minute from now a truck could very well go by me on the left.

I view the GFS's mid-July hurricane like that. There is a hurricane far distant down the GFS road. Does that mean a hurricane will strike the Carolinas then? No. It could go out to sea, strike Florida, or simply disappear. Or suppose the hurricane is shown going out to sea. Does that mean there will be a hurricane and it will spin fish? No. It could hit the US East Coast instead. But still the appearance of a storm on the GFS in mid-July, especially when it appears in a succession of runs, indicates that there is a possibility of a storm affecting somewhere around the Atlantic at that time. So I will keep watching it, especially if it is still there when it passes the 168 hour mark.

[vacanechaser]

I view GFS and other models like driving down a highway, especially one where you can see a long distance. I look down the highway. I see a tractor-trailer way way off as a small dot on the horizon. It is in the other lane. Does this mean it will pass by me on the road? No. It could turn right onto an intersecting road before it gets to where I am, or it could pull over at a rest stop and disappear from the highway. But still the appearance of this truck indicates that about a minute from now a truck could very well go by me on the left.

I view the GFS's mid-July hurricane like that. There is a hurricane far distant down the GFS road. Does that mean a hurricane will strike the Carolinas then? No. It could go out to sea, strike Florida, or simply disappear. Or suppose the hurricane is shown going out to sea. Does that mean there will be a hurricane and it will spin fish? No. It could hit the US East Coast instead. But still the appearance of a storm on the GFS in mid-July, especially when it appears in a succession of runs, indicates that there is a possibility of a storm affecting somewhere around the Atlantic at that time. So I will keep watching it, especially if it is still there when it passes the 168 hour mark.

Not to mention that the Euro is starting to show it as well.. Will be very interesting in a few days.. The water is plenty warm.. Brings back memories of 1996 and Bertha..

Jesse V. Bass III
http://www.vastormphoto.com
Hurricane Intercept Research Team

[cycloneye]

DISCUSSION...GOOD INSTABILITY AND RESIDUAL MOISTURE FROM A THETA-ERIDGE NORTH OF THE AREA ALLOWED THUNDERSTORMS TO GROW UP TO 60THOUSAND FEET THIS AFTERNOON WITH NUMEROUS AREAS INCLUDING SANSEBASTIAN...SAN JUAN...MOROVIS...AND CIALES THAT HAD AREAS NEAR OREXCEEDING TWO INCHES OF RAIN. THE THETA-E RIDGE WILL MOVE QUICKLYNORTH OF THE AREA AND A THETA-E TROUGH WILL CROSS PUERTO RICO ONWEDNESDAY AFTERNOON. A SECOND RIDGE ASSOCIATED WITH THE TROPICALWAVE WILL CROSS OVER THE VIRGIN ISLANDS AND PUERTO RICO ON WEDNESDAYNIGHT...WITH PRECIPITATION LINGERING THROUGH THURSDAY MORNING.ALTHOUGH THE EULERIAN MASS MODEL WANTS TO BRING IN CONSIDERABLEPRECIPITATION TO SOUTHEAST PUERTO RICO WEDNESDAY MORNING WITH ALMOSTTWO INCHES OF RAIN BETWEEN THERE AND SAINT CROIX...BELIEVE THAT THEGFS WILL HAVE A BETTER HANDLE ON THE WAVE. THE WAVE IS STILL TOO FARBACK BEHIND THE WINDWARD ISLANDS AND DRIER AIR IS ACCOMPANYING THEAREA AHEAD OF THE WAVE. NEVERTHELESS...THERE IS SOME CYCLONICTURNING IN THE 1000-850 MB LAYER WINDS WEDNESDAY MORNING THATAPPEARS LIKE A WEAK TROUGH MOVING THROUGH.DRIER AIR BEGINS INTRUDING ON THE AREA THURSDAY NIGHT AND CONTINUESTHROUGH SUNDAY MORNING. ANOTHER WAVE SHOWS GOOD CYCLONIC TURNING INTHE MODEL WITH A BAND OF EXCELLENT MOISTURE THAT MOVES THROUGH ONMONDAY. AFTER THIS THE GFS WANTS TO SPIN UP THE WAVE NOW LEAVING THEAFRICAN COAST INTO A LOW THAT CROSSES THE LESSER ANTILLES ONWEDNESDAY OF NEXT WEEK AND MOVES JUST NORTH OF PUERTO RICO ONTHURSDAY. THIS COULD BE A SPURIOUS LOW AND WILL NOT INCLUDE IT INANY GRIDS FOR NOW. BUT THE WAVE THAT MOVES THROUGH COULD PROVIDEMORE THUNDERSTORMS AND HEAVY RAIN FOR THE LOCAL AREA.

For the first time the San Juan office talks about what the GFS is showing for days.

[jimvb]

Did someone mention Bertha? I was biting my fingernails on that one. It struck NC and VA coasts right at the beginning of SUUSI! I was wondering if it would affect us getting there or other people that come from those areas. Fortunately, it produced only a little bit of rain and almost no wind, so it did not affect that conference too much.

[Wthrman13]

a note about the 6 and 18Z forecasts

the initial conditions are merely extrapolations of the previous 6 hour forecast. Basically, the analysis is the 6 hour forecast from before. They really shouldn't be used as sole forecast indicators, only the 0 and 12Z when the observations are included in the analysis

That's not entirely true. Available observations are indeed analyzed at the 06 and 18 Z times. There just aren't as many observations as there are at the synoptic times of 00Z and 12Z.

To expand just a bit...

Basically, the 0z and 12z models get all the new observations, and the 6z and 18z models are updates with the features moved along 6 hours then initialized. Basically the 6/18z runs are reruns with interoplated data.

Sometimes, but not often, some new obvs will get into the models at 18z. Every now and then the SPC will request special soundings for anticipated severe events, or the NOAA dropwindsondes will work their way in, etc. But by and large, there are really only 2 runs per day with a couple of updates at 2am and 2pm respectivelly.

This is why, generally, any big changes in forecast thinking occur in the 5pm and 5am advisory packages and not many (usually anyway) big model-dependant changes occur in the 11am/pm discussions.

MW

Actually, plenty of data is included in the 06Z and 18Z analyses. Again, just not as much, particularly RAOB data.

Dan

[artist]

this is interesting -
http://stormcarib.com/reports/2005/stlucia.shtml

Some weather coming?
From: "Marcus Day" <daym at candw.lc>
Date: Tue, 28 Jun 2005 06:31:40 -0400

Looks like we may have some weather in the Windwards in a day or two. It also looks like this could be one of those systems that strengthens when it passes into the Caribbean.

W TROPICAL ATLC TROPICAL WAVE IS ALONG 48W S OF 15N MOVING W
10-15 KT. WELL DEFINED LOW LEVEL CIRCULATION IS OBSERVED
ESPECIALLY ON THE LAST VISIBLE SATELLITE IMAGES. SCATTERED
MODERATE/ISOLATED STRONG CONVECTION IS W OF THE WAVE AXIS WITHIN
75 NM OF 12N FROM 49W-53W.


Marcus Day, D.Sc.
Director
Caribbean Drug Abuse Research Institute

--------------------------------------------------------------------------------

[cycloneye]

http://www.nco.ncep.noaa.gov/pmb/nwprod ... loop.shtml

00z loop.In this 00z run GFS downplays what it had in earlier runs.Only has 2 lows forming in the Tropical Atlantic and both dont do too much however both are strong waves.

[jimvb]

It seems that this "hurricane" has vanished. In the 00Z run, the Bermuda High is farther east and maintains a ridge to the SE US, causing the hurricane to hit southern Florida. In the 06Z run, I can't find it. It seems to dissolve into the milieu of the ITCZ. There is another storm in a few days that turns away from land early, but I don't think this is the same storm. Just like last year's SUUSI hurricane, this one has suddenly gone.

[clfenwi]

It seems that this "hurricane" has vanished. In the 00Z run, the Bermuda High is farther east and maintains a ridge to the SE US, causing the hurricane to hit southern Florida. In the 06Z run, I can't find it. It seems to dissolve into the milieu of the ITCZ. There is another storm in a few days that turns away from land early, but I don't think this is the same storm. Just like last year's SUUSI hurricane, this one has suddenly gone.

The model isn't developing it into a hurricane anymore, but it still has a low forming.

See http://moe.met.fsu.edu/cyclonephase/gfs/fcst/index.html (Cyclones 41 & 65 on the current imge). On the left you can see the cyclones the model was spinning up on previous runs. It certainly appears to me that the storm you saw turning away from land early is the same system it was developing into a hurricane in previous runs.

Looking at the same source, the CMC is forming something in the same position, but is not eroding the ridge the way the GFS is (thereby not turning the storm to open waters in the 144 hour period).

[Derek Ortt]

how many obs make it until the off analysis? There are very few US NWs station obs that make it in. I know some G-IV drops do make it in (which causes major model problems as it creates an asymmetry in the 18Z and then the 0Z)

[cycloneye]

http://www.nco.ncep.noaa.gov/pmb/nwprod ... loop.shtml

12z run does not show anything spectacular in terms of developing into a storm.But it has some weak lows some call them spurious moving across the Atlantic.

[vacanechaser]

http://www.nco.ncep.noaa.gov/pmb/nwprod/analysis/carib/gfs/12/index_ten_l_loop.shtml

12z run does not show anything spectacular in terms of developing into a storm.But it has some weak lows some call them spurious moving across the Atlantic.

Typical GFS... Same old song and dance.. Did this with Isabel too back in '03... Out of the gate, it was great..... nailed the forecast as far as landfall 14-16 days out... But after that, it didnt develop, it did, then into the Gulf hitting Mexico then out to sea... Euro and a few others are hinting at it now... Gonna get interesting..


Jesse V. Bass III
http://www.vastormphoto.com
Hurricane Intercept Research Team

[Wthrman13]

how many obs make it until the off analysis? There are very few US NWs station obs that make it in. I know some G-IV drops do make it in (which causes major model problems as it creates an asymmetry in the 18Z and then the 0Z)

I am not entirely certain on how many of each type of obs actually make it into the 18Z and 06Z runs. This website has a list of all the observations that are currently assimilated into the GFS model:

http://www.emc.ncep.noaa.gov/gmb/gdas/

Most of the data listed is available throughout the day. About the only data that isn't available routinely are the upper-air RAOB sounding data and dropsondes. Pretty much everything else should be available at 18Z and 06Z for assimilation into the initial conditions of the model. It would make absolutely no sense to have all that data available and to not use it in the model. If the 18Z and 06Z initial conditions for the GFS were simply extrapolations of the 12Z and 00Z runs, there would be no point in running them.

This web page has links to maps showing various obs used or rejected in the different model cycles (although curiously the surface obs don't appear to be getting into the model at any of the forecast cycles, including the 00Z and 12Z runs, so I don't know how accurate this page is). However, if you look at some of the recent 18Z and 06Z runs, you will see lots of ocean surface, satellite data, etc. getting into the model analyses.

http://wwwt.emc.ncep.noaa.gov/gmb/ssaha/

This said, it's true that the lack of upper-air soundings means that much of the model initial state at 06Z and 18Z is not appreciably modified from the previous 6 hr forecast from 0Z and 12Z, respectively, since soundings give you information about the entire depth of the atmosphere, and other data types are more limited. However, the data that is used does still have an impact.

Hope this clarifies some things,

Dan

[Derek Ortt]

one other interesting fact about the beloved GFS model,

it was discovered after Hurricane Ivan that <b>INNER CORE</b> dropsondes were getting into the model (yes, you heard me correctly, the mdoel with a 45km resolution was assimilating inner core sondes, creating nice gravity waves and the like which could really mess up a cane forecast). This problem is supposed to be corrected this year, but not for a few more weeks

[jimvb]

I see now on the GFS that the storm we have been talking about will go out to sea around July 7 and then dissipate. But then I notice another one that strikes southern Florida on July 13, then creeps up the west coast and strikes Appalachicola on July 16. If this happens, it will disrupt SUUSI, as many people who go there live in Florida. The path it will follow, according to GFS, seems to be halfway between Frances/Jeanne and Ivan. This is 16 days out, so it just appeared on the prediction horizon. I will wait to see if future runs sustain this, especially to within 7 days, when the GFS starts to become more accurate.

Incidentally, from another thread, I found out about a URL

http://radio.nhcwx.com/flghchp/displayimage.php?album=2&pos=12

describing a weather report that suggests that Bret was a dangerous storm - it could have developed to category 3 or higher if it had stayed out at sea. The weatherman at this video did mention something developing off the African coast.

[Scorpion]

Hmm interesting. Can you imagine a CV system hitting Florida in July?? Just think about what it will be like in September.